Artificial Agent Systems and Machine Learning Approaches to Studying Conflict and Political Violence

Abstract

In this Panel, we aim at bringing together research in various areas of political conflict and political culture which are currently employing new interdisciplinary methodologies. Some areas of core social and political culture research approached in this panel concern issues like friends and enemies in a small world, complex territorial approaches to state dynamics, and the relationship between social conflict and community identity. The papers presented in this panel share the approach of individual agents interacting in either small worlds or in large virtual worlds as groups, societies or states (polities). Approaches employ research methodologies based on the advanced technologies of the artificial, like agent-based systems, and advanced technologies of integrating data from complex sensors, like GIS-systems. As opposed to the limited power of data processing in the empirical (survey research) paradigm, these approaches reveal the power of such technologies to dynamically process huge amounts of data and extract patterns of preferences and their relationships. An agent-based system could thus model the dynamics of the relationship between data about the individual characteristics (vanity) and opinion propagation in a small world of individual agents. Agent-based models of territorial (border) dynamics could tentatively identify the drivers of colonial expansion and explain the state evolution from emergence to failure. The Panel aims at proving that ongoing research concerning political conflict as related to political culture phenomenology employs intensively the advanced technologies associated with the sciences of the artificial and develop toward new interdisciplinary approaches of the classic research issues, like polity, public opinion propagation, social influence, social and political conflict, and war.